Christian Marc Schmidt

Founding Partner @schemadesign. Exploring the edge of AI, spatial computing, & data. Formerly @pentagram, @ideo, @microsoft. Author of

[Christian Marc Schmidt @cms_]

We’re building the OS for creative teams. Memory you own. Alpha is open.

[Studiograph @studiographai]

We’re building Studiograph: the operating system for creative teams.

One place for everything your team makes and knows.
AI grounded in how you already work.

Private alpha is open, free while we shape it.
https://t.co/BdZbdDYn9M

[Christian Marc Schmidt @cms_]

What does it mean to be an AI-first consultancy?

We’ve been asking ourselves this question at Schema.

Looking for some insight.

[Christian Marc Schmidt @cms_]

@hwchase17 Single player, permissions are trivial. The agent acts as you.

Multiplayer, every action needs an answer to the question: on whose behalf, and with what scope? Who can read the brief, who can change a decision, who can close a task.

[Christian Marc Schmidt @cms_]

A client emails. An invite lands. A deal comes up in a thread.

Inbox turns each into tasks and records for the right project, person, and organization, with no retyping, and suggests deals.

Private alpha: https://t.co/HEAlV44QsC

[Studiograph @studiographai]

Work arrives by email and calendar invite. Inbox, in Studiograph, turns it into tasks and records on the right project, person, and organization, and suggests deals. Nobody needs to copy and paste. Request alpha access: https://t.co/rTM76rRXpU

[Christian Marc Schmidt @cms_]

Chat is where work gets discussed, not where it lives.

For long projects the AI should work on the shared artifacts: the brief, the decisions, the open tasks, with permissions for team collaboration.

It’s the difference between memory and knowledge. That’s our bet with @studiographai.

[Christian Marc Schmidt @cms_]

@boardyai @hf0 30

[Christian Marc Schmidt @cms_]

@studiographai now connects to Claude Code, Codex and Cursor. Your team can now use their existing subscription to work from your organization's projects and tasks.

Request alpha access at https://t.co/HEAlV44QsC

[Christian Marc Schmidt @cms_]

@illscience Just-in-time software, yes. But, people trust interfaces they've learned, and you can't learn interfaces that are different every time.

Generative interfaces will still be opinionated, built on rules about what stays fixed and what is variable.

[Christian Marc Schmidt @cms_]

The primitives will converge, but I'm less sure there will be a single OS.

What makes a person or team distinct is process and know-how, and that's what harnesses end up encoding. GStack is a good example.

I think we will see a long tail of harnesses, tailored to the specific needs of people and organizations.

[Christian Marc Schmidt @cms_]

@brexton @NotionHQ Agree on multiplayer. The open question is whether one general-purpose workspace fits every team.

Memory gets a lot more useful when it's opinionated about the work, who the client is, what was decided, what's next.

[Christian Marc Schmidt @cms_]

We ran our team on Notion for years, then took it one step further.

Our shared memory is plain markdown with typed entities: people, orgs, projects, decisions. Every agent reads the same structure, and the data stays truly portable, not locked in one app's database.

That’s why we built @studiographai.

[Christian Marc Schmidt @cms_]

The model is swappable.

What you can't afford to lose is the memory, what it knows about you, your work, and your team.

Context is the real moat, and every agent has a vested interest in keeping you locked in.

Keep it portable, own your data, and switch models or agents based on the task.

[Christian Marc Schmidt @cms_]

@thenanyu Hard to disagree given the current state of software. That said, we are about to see radically new interface paradigms emerge that will define the AI era. These will be simpler, conversational, and adaptive. No or minimal chrome. Voice as input, GUI focused entirely on artifacts.

[Christian Marc Schmidt @cms_]

@royrubin05 This is the value of the Obsidian to Claude Code flywheel and why we built @studiographai for teams. The frontiers are trying to lock you into their ecosystems. Data portability is key at the current pace of model innovation.

[Christian Marc Schmidt @cms_]

@garrytan Yes. And I'd go one step further - figuring out what this looks like for teams is the real opportunity. It's also a fundamentally different problem as it turns out.

[Christian Marc Schmidt @cms_]

@boardyai The graph view emerges as the team adds context to the workspace: anything from presentations to design systems, apps, proposals, and case studies. We are running our entire design studio from it — currently looking for alpha testers.

[Christian Marc Schmidt @cms_]

Meetings are core context for teams. We have been building Studiograph's Meeting Capture feature over several months. It writes straight into your graph, translating calls into the people, organizations, and tasks the meeting was actually about.

[Studiograph @studiographai]

Meeting Capture, a new feature in Studiograph, writes what happens in a call straight into your workspace, turning it into the people, organizations, and tasks it's about. You can also ask it questions live from the meeting chat.

[Christian Marc Schmidt @cms_]

@garrytan @AsideAI I've been liking @AsideAI quite a bit due to its agent friendliness. Might take over as my daily driver.

[Christian Marc Schmidt @cms_]

Private alpha is free while we shape the product: https://t.co/HEAlV44iD4

[Christian Marc Schmidt @cms_]

We've been busy onboarding private alpha teams and getting real feedback from the people using Studiograph day to day. We now have some big updates coming on Monday. Private alpha is still open if you want to see them land in your own workspace first.